<!DOCTYPE HTML>
<html>
<head>
<meta charset="UTF-8">
<title>Exercise 5.33: Parametrized l1-norm approximation</title>
<link rel="canonical" href="/Users/mcgrant/Repos/CVX/examples/cvxbook/Ch05_duality/html/ex_5_33.html">
<link rel="stylesheet" href="../../../examples.css" type="text/css">
</head>
<body>
<div id="header">
<h1>Exercise 5.33: Parametrized l1-norm approximation</h1>
Jump to:&nbsp;&nbsp;&nbsp;&nbsp;
<a href="#source">Source code</a>&nbsp;&nbsp;&nbsp;&nbsp;
<a href="#output">Text output</a>
&nbsp;&nbsp;&nbsp;&nbsp;
<a href="#plots">Plots</a>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="../../../index.html">Library index</a>
</div>
<div id="content">
<a id="source"></a>
<pre class="codeinput">
<span class="comment">% Boyd &amp; Vandenberghe "Convex Optimization"</span>
<span class="comment">% Jo&euml;lle Skaf - 08/29/05</span>
<span class="comment">% (a figure is generated)</span>
<span class="comment">%</span>
<span class="comment">% Let p_star(epsilon) be the optimal value of the following problem:</span>
<span class="comment">%               minimize    ||Ax + b + epsilon*d||_1</span>
<span class="comment">% Plots p_star(epsilon) versus epsilon and demonstrates the fact that it's</span>
<span class="comment">% affine on an interval that includes epsilon = 0.</span>

<span class="comment">% Input data</span>
A = [-2  7  1; <span class="keyword">...</span>
     -5 -1  3; <span class="keyword">...</span>
     -7  3 -5; <span class="keyword">...</span>
     -1  4 -4; <span class="keyword">...</span>
      1  5  5; <span class="keyword">...</span>
      2 -5 -1];
b = [-4 3 9 0 -11 5]';
d = [-10 -13 -27 -10 -7 14]';
epsilon = [-1:0.05:1];
p_star = zeros(size(epsilon));

fprintf(1,<span class="string">'Computing p*(epsilon) for -1 &lt;= epsilon &lt;= 1 ...'</span>);

<span class="keyword">for</span> i=1:length(epsilon)
    cvx_begin <span class="string">quiet</span>
        variable <span class="string">x(3)</span>;
        minimize ( norm( A*x + b + epsilon(i)*d, 1) )
    cvx_end
    p_star(i)= cvx_optval;
<span class="keyword">end</span>

fprintf(1,<span class="string">'Done! \n'</span>);

<span class="comment">% Plots</span>
plot(epsilon, p_star)
line([-.2 -.2], [2 14], <span class="string">'LineStyle'</span>, <span class="string">'--'</span>)
line([.5 .5], [2 14], <span class="string">'LineStyle'</span>, <span class="string">'--'</span>)
xlabel(<span class="string">'\epsilon'</span>);
ylabel(<span class="string">'p^*(\epsilon)'</span>);
title(<span class="string">'p^*(\epsilon) vs \epsilon'</span>);
</pre>
<a id="output"></a>
<pre class="codeoutput">
Computing p*(epsilon) for -1 &lt;= epsilon &lt;= 1 ...Done! 
</pre>
<a id="plots"></a>
<div id="plotoutput">
<img src="ex_5_33__01.png" alt=""> 
</div>
</div>
</body>
</html>